Optimizing the forex trading system parameters: AUD/USD

User Rating: / 2
PoorBest 
Written by Forex Automaton   
Thursday, 14 May 2009 14:12
Article Index
Optimizing the forex trading system parameters: AUD/USD
Optimizing the forecasting parameter
Optimizing the stop-loss parameter
Optimizing the trade entry parameter
Optimizing the trade exit parameter
Summary of progress
All Pages

This is the fifth report in the series of the buy-side forex trading system optimization reports for the individual currency pairs, traded on the day scale, which began with AUD/JPY. In the algorithm, the forecast signal whose nature will not be disclosed is fed into the money management framework driven by three adjustable parameters. The set of 13398 parameter combinations represents the totality of possible trading styles under study. The goal is to optimize the trading style by finding, on the basis of the simulated trading performance, such values of parameters as to maximize the return while minimizing risk. The insights obtained in the process may be of general interest, since the problem is common to all traders, robots and humans alike.

Analysis approach and the data set

It can not be overemphasized that for the analysis to be of any value, the algorithm may not trade the data used to train its decision making. The algorithmic learning is continuous and lasts as long as the trading lasts, but is limited to the past data. A run of the program included simulations of trading histories of over 13,000 independent "virtual traders" (forex robots), each of them being an incarnation of the same algorithm, differing by the setting of the adjustable knobs. This report uses the AUD/USD day scale data (daily open, close, low and high) covering the time interval from August 20, 2002 to March 23, 2009. The key concepts of conditional projection distributions and profile histograms have been explained before.

Going through a similar study for every major forex exchange rate involving USD remains my plan, with USD/CHF and USD/CAD remaining to be covered. The reader is encouraged to go through similar reports for other currency pairs. In the process, many things related to money management in the context of such an algorithm are becoming clear, and I am zooming down on a "golden" patch in the money management parameter space.

In the production regime, all major forex exchange rates will be analyzed simultaneously, therefore the dimensionality of the picture and the computational complexity of the "AI" procedure will grow considerably. So will the CPU demands of the optimization procedure. But the money management insights extracted from optimizing the individual currency pairs will hopefully prove helpful in limiting the multitude of virtual traders to simulate, which will speed up the optimization.

Among the currency pairs looked at so far, significant commonalities as well as differences have been seen in the way performace responds to the changes in parameters. While interpretable commonality is good news, too much eclectic variety in response may indicate that either the variables or the way they are looked at need re-thinking.

Consistent performance with consistent selection of parameters is what I am after, and the picture indeed begins to look like this is the case, although this will be the subject of a separate analysis.



Last Updated ( Monday, 04 January 2010 12:35 )